Among users of cochlear implants or middle ear implants there is a strong desire for a small, fully implanted system. Any implanted microphone for such a system must detect external sounds after attenuation by a layer of skin, but must discriminate between these sounds and unwanted vibrations of biological origin, such as the user's voice. In this research we will optimize parameters for, and demonstrate the effectiveness of two microphone architectures, referred to as the "suspension microphone" and the "acceleration microphone." Preliminary data is presented showing the advantages of each in discriminating between sound and tissue-borne vibration. Our specific aims are to: 1. Optimize vibration sensitivity in a suspension microphone. We will construct suspension microphones with physical parameters expected to result in good isolation of acceleration pressures. We will measure the response of these microphones, while under a layer of simulated skin, to air-conducted sound and to vibration, then modify the design as indicated by comparison of the results to prediction. 2. Optimize vibration sensitivity in an acceleration microphone. We will construct acceleration microphones, model, test and optimize the design as for the suspension microphones. 3. Demonstrate microphone performance in cadavers. Optimized suspension and acceleration microphones will be used for confirmatory testing in cadavers. We will prepare specimens with bone beds at sites suitable for implanted hearing system microphones. Sound and vibration stimuli will be applied, responses measured, and performance predicted for a fully implanted hearing system.